IBM 1620

[Chuck Guzis]

On 08/28/2015 05:59 AM, Paul Koning wrote:

> Especially if you didn't have the "direct seek" option.  I had a
> one-card program that would do for (i=0; i<100; i++)
> seek_cylinder(i).  That doesn't step track to track as you would
> expect; instead, it goes from track n to 0 to n+1, resulting in
> progressively wilder and slower seeks as n increases.

Yup, the one I worked with did not have the "direct seek option".  What 
was interesting from a historical standpoint was the way in which the 
1311 was used by Monitor II-D.   IIRC, the first 25 cylinders were 
dubbed "work cylinders" with no formal filesystem.  The remainder was 
used for more-or-less permanent storage of programs--not so much as a 
data store.

The CADET I worked on did have the Indirect Addressing feature, which I 
believe was required to run Monitor II-D.  Hard today to think of an 
addressing mode as an optional ($$$) feature.  The Model II had an index 
register.

I can still remember many of the 1620 numeric opcodes.  Strange, 
considering that I can't remember where I put my keys or what I had for 
breakfast...

--Chuck